Ferrostaal Corp. v. MV Singa Wilguard, 838 F. Supp. 757 (S.D.N.Y. 1993)
November 12, 1993
United States District Court, S.D. New York.
*758 Healy & Baillie, New York City (John C. Koster and Evanthia Coffee, of counsel), for plaintiff.
Mahoney & Keane, New York City (Cornelius A. Mahoney, of counsel), for defendants Wilguard Shipping Corp. and Anders Wilhelmsen & Co.
Snow Becker Krauss P.C., New York City (Alan Van Praag, of counsel), for defendant Atlantic Lines, S.A.*759 OPINION, FINDINGS OF FACT AND CONCLUSIONS OF LAW
NEWMAN, Senior Judge:
Ferrostaal Corporation, the consignee of a shipment of hot rolled steel rail from Duisburg, Germany, brings this action against defendants pursuant to the Carriage of Goods by Sea Act ("COGSA"), 46 U.S.C.App. §§ 1300 et seq. (1988). The cargo had been ordered from Ferrostaal by the New York City Transit Authority ("NYCTA"), and was shipped aboard the M/V SINGA WILGUARD, departing Antwerp, Belgium on September 16, 1989 and arriving at Newark, New Jersey on October 5, 1989. Ferrostaal seeks to recover $159,680.15, plus interest and costs, representing the expenses and loss incurred in connection with the rejection of the rail by NYCTA and subsequent salvage sale to the Port Authority Trans-Hudson ("PATH").CONTENTIONS OF THE PARTIES
Initially, it is uncontested that hot rolled steel products will form a surface layer of atmospheric rust following exposure to ambient humidity or rain, and such corrosion does not damage the steel. The chlorides present in seawater, by contrast, are aggressive agents that penetrate below the surface and will proceed to destroy the steel unless removed soon after contamination.
Ferrostaal asserts that the WILGUARD was unseaworthy due to defendants' failure to maintain a watertight hatch. It is contended that seawater entered through the hatch cover and poured into hold # 1 on the WILGUARD during a storm in the North Atlantic, thereby contaminating the stacked cargo of rails. Ferrostaal's claim is predicated entirely upon the contention that the rail suffered from seawater corrosion and pitting.
Defendants concede that upon discharge the rail was covered with a layer of atmospheric rust, but argue that the cargo was in fact undamaged for the purposes of its intended use. Although NYCTA cited the rusty condition of the rail as the reason for rejection, defendants urge that the rail was entirely usable and that the true reason for NYCTA's reluctance to accept the rail stemmed from unrelated commercial disputes between Ferrostaal and NYCTA.
The court has admiralty and maritime jurisdiction of this matter pursuant to 28 U.S.C. § 1333 (1988). A bench trial was conducted by the writer from April 19 to 22, 1993. The following constitutes the court's findings of fact and conclusions of law in accordance with Fed.R.Civ.P. 52(a).THE RECORD
The record in this case consists of, among other things, the transcript of the live testimony of five witnesses for Ferrostaal and six witnesses for defendants, as well as 407 exhibits. Additionally, the parties referred to the deposition transcripts of several persons who were not produced at trial.
Ferrostaal presented the testimony of the following five witnesses: Dietrich Seraphin, an officer of Ferrostaal; Hugh Fowley, a surveyor who observed some of the rail at Newark; Severo Calima, an employee of I.S. Derrick Independent Ship and Cargo Surveying, Inc. who examined the tank tops and other cargo after the rail was discharged; Dr. Klaus Wick, a metallurgist employed by Thyssen, Inc., which manufactured the rail for Ferrostaal; and Capt. John Alder, as an expert on marine surveying.
Defendants produced the following six witnesses: George Farrell, a surveyor who tested the cargo at Providence, R.I. and Newark, N.J.; J.W. Laudermilch, president of Steel Services Company, Inc.; Capt. Arthur Sparks, a leading authority on marine surveying; Rey Ortiz, a surveyor who tested the cargo at Newark on behalf of Ferrostaal; Dr. John Francis Molina, a chemist at New York Testing Laboratories; and Howard Sabin, a corrosion engineer with impressive credentials, to whose testimony the court gave considerable weight.*760 FINDINGS OF FACT
Ferrostaal Corporation is organized and existing under the laws of the State of New York and is the American subsidiary of Ferrostaal AG, a large German trading company. Ferrostaal is engaged in, among other things, the import and export of rail and railroad products. At all times relevant to this action, defendants Wilguard Shipping Corp. and Anders Wilhelmsen & Co. were foreign corporations with offices and a place of business in Oslo, Norway, and were the owners of the WILGUARD. Defendant Atlantic Lines, S.A., with an office and place of business in New Jersey, was the charterer of the WILGUARD during the voyage. All defendants were engaged in business as common carriers of merchandise by water for hire.
The rail which forms the basis of the instant case was manufactured at Thyssen's plant in Duisburg, Germany pursuant to an order placed by NYCTA with Ferrostaal in December 1988. The rail was covered with mill scale, an oxide that forms as a natural and inevitable by-product of the manufacturing process of any hot rolled steel product. After manufacture, the subject rail was not shipped immediately; rather, it was stored outside Thyssen's plant for approximately two months due to a series of disputes between Ferrostaal and NYCTA relating to Thyssen's noncompliance with NYCTA's manufacturing specifications. During this period, the rail was stored outdoors on the piers of the mill port in Duisburg, where it was exposed to the weather.
Duisburg is located in the Ruhr industrial region of Germany, within 40 miles of Dusseldorf. Although Germany has strong restrictions on industrial pollutants, the area is heavily industrialized, and certain amounts of sulfur emissions are to be found due to the presence of a coke battery in Duisburg. All of Northern Europe is susceptible, moreover, to acid rain, which can accelerate the process of normal atmospheric corrosion on hot rolled steel surfaces.
The problems associated first with the manufacture and then the shipment of the rail appear to have been legion. Before the rails were even loaded on board the WILGUARD, Dietrich Seraphin, an officer of Ferrostaal who was in charge of the order, wrote to Ferrostaal's German parent in Essen, "[n]othing seems to go right with this order." Defendants' Exhibit "D." Initially, Ferrostaal failed to deliver the rail to NYCTA on time, prompting a threat by NYCTA that it would cancel the order. More important, however, was the failure of Thyssen and Ferrostaal to provide rail according to NYCTA's specifications. Seraphin explained that NYCTA's requirements differed from those of most other urban mass transit systems in a number of respects, particularly with regard to the size and composition of its tracks.
Specifically, NYCTA required 100 lb. rails, a portion of which were to be drilled. Some of the rail was not drilled. The rails were to be "end hardened," apparently in order to allow the rail to withstand the heavy stress of subway trains running over the connections. Thyssen, however, did not end harden the rail. Additionally, the manganese content of the Thyssen product exceeded that of 100 lb. rail. The specifications further called for the rail to be chamfered; the rail was not chamfered, although NYCTA later waived that specification. Ferrostaal negotiated discounts and ultimately persuaded NYCTA to accept shipment of the nonconforming goods.
Following NYCTA's agreement to take delivery, the rail was transported in open barges from Duisburg to Antwerp. The trip to Antwerp took 24 hours. On the piers in Antwerp, the rail was re-sorted and loaded into two different cargo holds. The cargo was shipped pursuant to two bills of lading, containing the clauses, "rust stained" and "wet before shipment," which were inserted pursuant to the Pre-Shipment Survey Report by the shipowner's surveyor, Sparks & Co., N.V. The survey report defines the relevant clauses as follows:*761 "Rust stained" Metal surface completely covered with a fine film of rust which has not resulted in deterioration, or impaired in any way the good condition of the cargo involved. "Wet before shipment" This clause is applied to all steel products, which show signs of wetness prior to shipment. This wetness is considered to promote a rust condition or accelerate the further development of any existing rust visible on the goods.
Plaintiff's Exhibit 53 at 3-4.
Photographic and testimonial evidence further confirm that upon loading at Antwerp the cargo was generally soaked by rainwater and was already covered with a uniform layer of light to moderate atmospheric rust. See, e.g., Plaintiff's Exhibit 253, Report of Capt. G. Van Leeuwen.
The WILGUARD departed Antwerp on September 16, 1989. During the transatlantic passage, the WILGUARD experienced storms measuring up to force 8 on the Beaufort Scale. The record demonstrates that ventilation of the ship's holds could not be undertaken for much of the voyage, and any attempt to do so would likely have resulted in the entry of seawater into the holds and damage to the cargo.
WILGUARD arrived at Brenton Reef, Rhode Island on or about September 29, 1989, where it remained at anchor for several days, during which time the holds were ventilated. The hatch covers were not opened at this time. See Deposition of Captain Joscelyn Colacao, Plaintiff's Exhibit 237, at 124, ln. 20, et seq.; Logbook, Defendants' Exhibit "XXXX" at September 29, 1989 to October 2, 1989. The WILGUARD arrived at Providence on October 3, 1989, where discharge operations took place on at least three of the hatches. At that time, the first of several stateside surveys of Ferrostaal's rail cargo took place.Inspection and testing.
The survey at Providence was conducted by George Farrell, who represented the shipowner's Protection and Indemnity Club and testified for defendants. As Farrell's Providence survey report indicated, heavy condensation was discovered within all of the hatches and on the cargo, apparently as a result of the closure of the hatches during the storms in the North Atlantic.
On October 3, 1989 Farrell boarded the vessel along with another surveyor. The logbook indicates that hatch # 2 had been opened on that date at or around 8:00 a.m. as part of discharge operations, and that hatches # 4 and 5 were opened subsequently. On arrival, Farrell asked that the # 1 hatch be opened for purposes of his survey. Farrell and Schell observed from the deck what they considered to be a "drip-down" pattern on the rails in the # 1 hatch. The record does not indicate whether a similar condition existed in the # 5 hatch. Farrell alone entered the # 1 hold, and conducted around twenty-five silver nitrate tests. Farrell testified:A: Well, you look for rust patterns. You look for, you know, drip down. Where the drips drip down, rust patterns occur. That's the area you want to test and *762 corner of the hatches where you sometimes see intrusions of water. And I walked over the cargo along the rust pattern and did random tests in sites I though might produce some results. Q: What were the results? A: They were negative. I couldn't find any seawater in there.
Record at 315-16 (emphasis added).
Ferrostaal places great weight upon a statement in Farrell's survey report in which he recites a belief that "some seawater intrusion" took place in the # 1 hatch. Plaintiff's Exh. 59 at 11. Ferrostaal urges the court to discount Farrell's contrary trial testimony and conclude that seawater contaminated the rails, only to be "leached out" by rain water subsequent to arrival at Providence.
Initially, the court considers that Farrell's testimony was entirely candid and that his statements were reasonably consistent. Indeed, the court was concerned to clarify this matter and elicited the following:Q: And you believed [Farrell's suspicion that seawater had intruded into the # 1 hold] to be true when you wrote it? A: I believed that there had been some seawater that had gotten into number 1, but I couldn't find any significant evidence of it.
COURT: What do you mean by some?A: When a hatch is opened on a vessel, there could be seawater trapped in the cover joints, and it can dump out into the coamings and dump out into the cargo. It could be a couple of cupfulls. If there is any major amount of seawater, it is going to show up dramatically on silver nitrate test. There is no trouble finding it. Number 1, I didn't, couldn't find it. Number 2, no problems.
Record at 351.
It remains, therefore, to address Ferrostaal's theory that saltwater present on the rail was "leached out" by rain falling into the hatch during the time that the hatches were opened, or by the drip-down of freshwater condensation in the hold. If such were the case, Farrell's test results at Providence would not deserve great weight.
Ferrostaal's theory is illogical, however, and it is defeated by the record. First, it is undisputed that significant amounts of seawater entered the # 2 hold during the voyage and contaminated a cargo of coils contained in the hold. Additionally, as noted supra, the evidence establishes the following: the hatches were not opened while the ship was at anchor at Brenton Reef; the # 2, 4 and 5 hatches, were, however, opened when the ship arrived at Providence; and the # 1 hatch was also opened on the same day at Farrell's direction. If, as Ferrostaal urges, freshwater had "leached" out chlorides from the surface of cargo in the # 1 hold, it would be unreasonable to expect that the cargo in the # 2 hold would have escaped the same effects. And yet Farrell found ample evidence of chlorides in hold # 2.
Ferrostaal presses further that hold # 1 experienced an ingress of up to 3 feet of standing seawater. For this proposition, Ferrostaal cites, inter alia, a portion of Farrell's survey report, which on closer examination relates only that the "[t]ank tops were wet with condensation with very occasional areas of standing water no deeper than ¼"." Plaintiff's Exhibit 67 at 5. There is no evidence that the water to which Farrell referred was seawater as opposed to puddles of rain or ship sweat.
Ferrostaal also offered the testimony of Severo Calima, who identified a dark spot on a photograph as a water line inside the hold. See Plaintiff's Exhibit 90. The court credits the contrary testimony of Capt. Sparks, and accordingly determines that no such quantity of water invaded the hold. Record at 449-57.
The WILGUARD proceeded to Port Newark, New Jersey, arriving on October 5, 1989. The rail was discharged by Maher Terminal on October 5 and 6, 1989.
At Port Newark, Farrell again entered the # 1 hatch to examine and conduct tests on the rail. By this point, the cargo was found to be dry. Farrell was accompanied by Richard Tischler, who acted on behalf of the Maher Terminal, which was hired by Ferrostaal to perform the discharge of the rail. Farrell and Tischler each conducted approximately *763 30 silver nitrate tests on the rail at Port Newark, most of which yielded negative results.
Farrell submitted samples of rust scrapings and dunnage to New York Testing Laboratories, Inc. for testing. New York Testing determined that the rust samples were free of chlorides altogether. The wood samples revealed traces of chlorides. Although the report (dated August 17, 1990) did not rule out seawater as the source, it also indicated the possibility of other causes.
Seraphin testified that he attended the discharge operations on Monday, October 5, 1989. He was accompanied by Rey Ortiz, a surveyor from Ewig International, which had been engaged by Ferrostaal. Seraphin and Ortiz entered hold # 1 and observed the cargo, which was covered with rust and indications of a drip-down pattern on the sides of certain rails. Seraphin did not enter hold # 5, but stated that he viewed the cargo from above deck, finding it to be covered with nothing more serious than light brown rust. Record at 21.
Ortiz' account agrees with that of Seraphin in some respects. However, Ortiz testified that the rail in hold # 1 was not different in appearance from the rail in hold # 5. Furthermore, according to Ortiz' account, Seraphin stated that the cargo "look[ed] fine" during the discharge. Record at 533.
Ortiz performed his survey for Ferrostaal from October 5 to October 6, taking between 50 and 60 silver nitrate tests on each day. Record at 531. All of the silver nitrate tests taken by Ortiz resulted in negative findings. Significantly, Ortiz' initial report, dated October 10, 1989, states that the rails were found with "light to moderate rust evenly," and recited the negative silver nitrate results. The message states that cargo on outturn was found "generally in good condition," save for six rails that were slightly bent. Ortiz' final outturn report, dated October 26, 1989, confirmed his initial findings, concluding that the cargo was lightly rusted due to ship sweat and condensation. The photographs appended to the report demonstrate that, with the exception of certain dark spots that were the result of either drip-down or contact with wet dunnage, the rail was covered by a uniform layer of rust. Plaintiff's Exhibit 81.
Ferrostaal argues strenuously that Ortiz' observations should be discounted, contending that Ortiz was an inattentive and inexperienced surveyor. The record demonstrates otherwise. Ortiz had been a marine surveyor for eight years prior to examining plaintiff's rail, during which period he surveyed approximately fifty or more steel cargos. The court further observes that at the time that Ortiz inspected Ferrostaal's cargo and subsequently prepared his survey report, he had been engaged by Ferrostaal. If the cargo in hold # 1 were to have been as noticeably different in appearance from the rail in hold # 5 as Seraphin suggested, it would be reasonable to assume that such fact would have been reported contemporaneously and would have appeared in the Ewig survey report. In view of the foregoing, the court finds that the rail from the two cargo holds was not different in appearance or condition at the time of discharge.
Before proceeding further, the court addresses the evidence given by Hugh Fowley, another witness for Ferrostaal. Fowley boarded the WILGUARD at Maher Terminal on October 5, 1989, in connection with a survey of steel plates that were due to be discharged at Philadelphia. Searching for the plates, Fowley entered hold # 1 and decided to conduct a gratuitous survey of Ferrostaal's rail. During this survey, he conducted silver nitrate tests on the rail, obtaining two "strong" positive results.
*764 Although a positive silver nitrate test would normally support an inference that chlorides such as are found in seawater contacted the rails, the court discounts the results of Fowley's tests for a number of reasons. First, Fowley stated that he mixed his own silver nitrate solution, rather than purchase the mixture from a chemist. The composition of his solution, however, did not conform to the Lloyds Survey Handbook method and would indicate positive results for alkalies. Second, Fowley did not take any rust samples for subsequent testing in a laboratory, as did the other surveyors in the case at bar. Since a silver nitrate test is only a test for chlorides, a laboratory report by a qualified chemist is necessary to confirm the presence of seawater. The court accordingly declines to draw an inference of seawater contamination based on Fowley's results, and discards his observations generally as to the condition of the rail.
On or about October 13, 1989, NYTCA's representative telephoned Seraphin from Brooklyn. NYCTA had received the first three truckloads of rail from the terminal, and expressed an intention not to receive any more, citing the rusty condition of the rail. The following week, NYCTA's representative met with Seraphin at Newark to examine the rail. It was determined that certain of the rail failed the "fish plate" test. Ferrostaal and NYCTA entered into a long series of negotiations whereby NYCTA agreed to accept part of the delivery. Two truckloads of the rejected rail were sandblasted; these rails then passed the fish plate test and were accepted by NYCTA.
Rather than obtain NYCTA's consent to sandblast the rest of the rail, Ferrostaal would ultimately sell the balance to A & K Railroad Material, Inc., which then sold the rail to PATH. Interestingly, J.W. Laudermilch, who testified on behalf of defendants, stated that all of the PATH rail passed the fishplate test, although it is not clear whether this resulted before or after sandblasting.
Significantly, Ferrostaal did not produce any representative of NYCTA, A & K Railroad, or PATH as a live trial witness.
Ferrostaal submitted the Ewig survey along with a claim notice to its cargo insurance company. The claim was delayed, however, because Ewig was not a surveyor approved by the underwriters. Ferrostaal thereupon appointed Luard & Company to conduct a second survey.
Luard visited the Maher Terminal and conducted an inspection of the rail three weeks after the date of discharge and took several photographs of the rail at that time. Although Luard reported heavy rust conditions on the rail, silver nitrate tests yielded negative results. On or about October 30, 1989, Luard sent rust scrapings to the laboratory of Stillwell & Gladding for analysis. The laboratory reported back to Luard that no evidence of seawater wetting was present.
Consequently, Luard's final report indicated that there was no evidence of seawater contamination, and Ferrostaal's insurance carrier declined to take further action on the claim.
NYCTA performed its own tests on samples of rust scale taken from the rail. NYCTA's laboratory report, dated January 19, 1990, stated, "sodium chloride was not detected." Defendants' Exhibit "TTT." Wick visited the terminal on behalf of Thyssen approximately one month later and took additional samples, which subsequently tested negative for seawater. Record at 253.
Long after its first report, Luard revised its assessment pursuant to the report of another surveyor whom Ferrostaal had engaged some eight months after the rail was discharged. The surveyor, William Bowes of *765 I.S. Derrick Independent Surveyors, did report positive silver nitrate tests from certain 8' by 8' timbers on which the rail was stacked at the Maher Terminal. The timbers belonged to the terminal and were not dunnage from the holds of the WILGUARD. Based upon those results, Luard then changed its survey report.
The court declines to follow Luard's example and places no reliance upon the Bowes report. Any probative value of Bowes' tests is immediately cast in doubt when one considers that they occurred several months after discharge. Furthermore, such positive results as were obtained were confined to areas of the rail that were tangent to the timbers; similar positive results were obtained from the lumber itself. Record at 82-3. The dubious weight of Bowes' silver nitrate results is further diminished in the light of laboratory testing by Dixie Services, Inc. Concerning the rust samples, the Dixie Services report of August 21, 1990 concluded that the rail had come into contact with either freshwater or substances "of other than seawater origin," probably calcium chloride. Rust scrapings taken by Bowes from the dunnage yielded inconclusive results, but indicated a foreign substance other than seawater. As to the timbers, the lab results disclaimed a conclusive finding of seawater contamination, suggesting only the possibility that the timbers had been contacted either by brackish water or seawater that had been washed away by freshwater.
Whatever value Luard may have attached to the Bowes report, it does not make the case for seawater; nor even if it did, could the report sustain a finding that the contamination took place in the holds of the WILGUARD. A careful reading of the Dixie Services reports leads to the conclusion that it is at least as likely as not that the timbers were contaminated by chlorides other than sodium chloride on the pier prior to the arrival of the rail, and that the timbers thus contaminated thereafter impregnated those portions of the rail that were stacked upon them. See Defendants' Exhibit "DD". It is likewise possible that chemical agents other than saltwater washed onto the rail and the lumber together, long after discharge and stacking. In either case, it is clearly impossible to determine, predicated upon results so attenuated in time from the date of outturn, that the rail was more likely contaminated while in the custody of the carrier than at a later date.
The remaining scientific evidence is of little value and may be disposed of quickly. Subsequent to Bowes' inspection, Ferrostaal engaged a metallurgist, Lucius Pitkin & Co., to examine the rail. The Pitkin report does not indicate seawater damage, and concludes that the corrosion was probably cosmetic, with no adverse structural consequences. See Defendants' Exhibit "OOOO." Importantly, several months separate the time of Pitkin's examination from the time of discharge. Another report by Robert Lyon & Co., engaged by Luard, made a qualified conclusion of seawater damage based upon Luard's photographs, which had been taken some three weeks after discharge. The Lyon report also left open the possibility of other causes such as freshwater and potash. The Lyon report was not based upon laboratory testing and is not accorded any weight by the court.
Finally, a series of tests conducted by Calima and Dixie Services were submitted into evidence by Ferrostaal; Calima's tests were performed after discharge at later ports, and concentrated on cargo other than the rail. Dixie Services also failed to give a conclusive pronouncement of seawater even as to these tests. See Plaintiff's Exhibit 87. The court similarly attaches no weight to this evidence.
In summary: the court observes that the cargo in the case at bar was the subject of exhaustive chemical and metallurgical examination over a period of more than twelve months. Virtually all of the results of such testing pronounced the rail to be free of chlorides. The only positive silver nitrate tests that were reported cannot be given any weight. Fowley's results are tainted by deviation from standard procedure and the failure to confirm his results through laboratory testing. Bowes' results are too attenuated in time from discharge and were unconfirmed by the Dixie Services report.
Additionally, in light of the considerable number of contrary results obtained by the *766 other surveyors, the validity of which is unchallenged, the court finds that the aggregate of silver nitrate tests taken in this case does not support a finding of seawater contamination to Ferrostaal's rail. More, the various laboratory reports either ruled out seawater altogether or provided no more than qualified statements as to the possible presence of either seawater or other contaminants. The short of the matter is that no single chemical analysis conclusively identified traces of seawater in any rust sample. Therefore, the court determines that the preponderance of the laboratory and metallurgical evidence much of which was generated for Ferrostaal and NYCTA decisively favors defendants.Expert testimony.
The ultimate issue of fact in the case at bar is whether upon discharge at Port Newark the cargo was in good condition or bad. The only remaining evidence to consider is the visual appearance of the rail itself, as demonstrated by photographic evidence taken close in time to the discharge of the rail, as well as the expert testimony.
Both sides offered a large number of photographs of the rail at each stage of the journey from Antwerp to Port Newark. The photographs establish plainly that the rail was covered with light to moderate atmospheric rust at loading, and that some of the rail had developed an accelerated stage of moderate and flakey surface rust at discharge. Ferrostaal places considerable emphasis on a number of photographs taken by Luard & Co. in particular, which photographs are appended to the Luard survey report. Plaintiff's Exhibit 56, photographs 5, 6, & 7. For the reasons that follow, the court determines that the photographs display no more than the effects of normal atmospheric corrosion, possibly accelerated by acid rain in Northern Europe, the ambient humidity in the cargo hold and whatever rainfall that took place during the three weeks of storage at the Maher Terminal that preceded the taking of the photographs.
The court found the testimony of Howard Sabin, the corrosion engineer, particularly helpful. Sabin acknowledged that the appearance of the rail in the photographs appended to the Luard report indicated severe atmospheric corrosion. However, Sabin refuted the conclusion of Ferrostaal's metallurgist, Klaus Wick, that the corrosion was the result of an aggressive agent, i.e., saltwater. Rather, the substance of Sabin's testimony was that the corrosion appearing on the WILGUARD's rail cargo was a combination of galvanic and ordinary atmospheric corrosion, accelerated due to the presence of mill scale, heavy condensation in the cargo hold and possibly acid rain. Record at 578, 603. It was Sabin's view that the rail was not damaged beneath the surface and that the rail could be easily returned to a commercially acceptable condition by wire brushing or sandblasting. Id. at 596.
The issue as to whether the rail was pitted was also sharply in contention. Pitting would reduce the "fatigue life" of the rail and render it unsuitable for prolonged service as subway track. To prove its contention, Ferrostaal offered first Fowley's testimony that pitting could be determined visually, by scraping the rail and running his hand over the surface of the rail. Dr. Wick stated that he could not see pitting, but extrapolated that the rail was pitted based upon the surface rust condition. Tellingly, however, none of Ferrostaal's witnesses had undertaken to determine the existence and depth of the alleged pitting scientifically, through the use of a pit gauge or of a micrograph, although such testing is the standard method for detecting pitting on steel surfaces.
Finally, Laudermilch testified on behalf of the carrier that he removed areas of corrosion by wire brush and inspected the metal in the same manner as Fowley; but Laudermilch observed no signs of pitting. Record at 376-77. Having considered and weighed the testimony, the court finds that Ferrostaal has not proved that the rail was pitted.
NYCTA had cited rust as the reason for rejection. Even accepting NYCTA's stated reason at face value, however, the evidence does not establish that the rail was damaged. The record shows that NYCTA accepted a portion of the rejected rail following sandblasting, although it was reluctant to authorize further sandblasting on the grounds that it was not in the market for refurbished rail. *767 It should further be recalled that NYCTA's own lab report yielded negative results for seawater contamination. Finally, the balance of the rail was later inspected for PATH, pronounced acceptable, sandblasted, and purchased by PATH for use in its system, thereby confirming Sabin's view of the true condition of the rail.
In view of the foregoing, the court finds that the rail was in good condition on outturn.CONCLUSIONS OF LAW I.
The court first addresses a threshold matter concerning Ferrostaal's failure to provide timely notice of damage to the carrier. The record establishes that Ferrostaal waited until three weeks after discharge to provide notice to defendants, which was two weeks after NYCTA had rejected the shipment. Defendants contend that Ferrostaal's claim is, therefore, defeated by operation of statute, namely, section 3(6) of the COGSA, which provides in relevant part:Unless notice of loss or damage and the general nature of such loss or damage be given in writing to the carrier or his agent at the port of discharge before or at the time of the removal of the goods into the custody of the person entitled to delivery thereof under the contract of carriage, such removal shall be prima facie evidence of the delivery by the carrier of the goods as described in the bill of lading. If the loss or damage is not apparent, the notice must be given within three days of the delivery.
46 U.S.C.App. § 1303(6) (1988).
As holder of the bills of lading, Ferrostaal was the "person entitled to delivery." Since the rust condition was readily apparent, and the possibility of saltwater contamination was discoverable at discharge through silver nitrate testing, Ferrostaal was required to give notice of the damage as of the time the rails were delivered to its custody. Consequently, the failure to give such notice for three weeks would defeat Ferrostaal's claim altogether, absent any contrary evidence concerning the condition of the cargo on outturn. See, e.g., M.W. Zack Metal Co. v. The S.S. Birmingham City, 291 F.2d 451, 453 (2d Cir. 1961); Leather's Best Intern., Inc. v. MV Lloyd Sergipe, 760 F. Supp. 301, 309-10 (S.D.N.Y.1991). Any such presumption of good condition on delivery falls, however, once the plaintiff adduces any credible evidence tending to show that the cargo was damaged prior to delivery. See Pacific Employers Ins. Co. v. M/V Gloria, 767 F.2d 229, 238 (5th Cir.1985). The record does, indeed, show that Ferrostaal has produced some evidence suggesting that the corrosion was more than superficial and took place while the rail was in the care, custody and control of the carrier. The prima facie defense of good delivery under section 3(6), therefore, does not end the court's inquiry.II.
That Ferrostaal has overcome the presumption of good delivery does not, of course, mean that Ferrostaal has necessarily prevailed on the ultimate issue by the preponderance of the evidence. To recover from defendants under COGSA, Ferrostaal must meet its burden of proof as to each of the two statutory elements of a prima facie case, namely, delivery of the cargo to the carrier in good order and condition and outturn in a short or damaged condition. See 46 U.S.C.App. §§ 1303(3), 1304 (1988). See also, Raphaely Intern. Inc. v. Waterman S.S. Corp., 972 F.2d 498, 501 (2d Cir.1992), cert. denied, ___ U.S. ___, 113 S. Ct. 1271, 122 L. Ed. 2d 666 (1993); Insurance Co. of N.Am. v. S.S. "Globe Nova", 820 F.2d 546, 548 (2d Cir.), cert. denied, 484 U.S. 965, 108 S. Ct. 454, 98 L. Ed. 2d 394 (1987); M. Golodetz Export Corp. v. S/S Lake Anja, 751 F.2d 1103, 1109 (2d Cir.), cert. denied, 471 U.S. 1117, 105 S. Ct. 2361, 86 L. Ed. 2d 261 (1985).III.
The court first considers Ferrostaal's initial burden of showing good condition *768 of the cargo at loading. It is fundamental that a bill of lading constitutes prima facie evidence of receipt by the carrier of cargo as it is described therein. 46 U.S.C.App. § 1303(4). Thus, a bill of lading containing no clauses or notations of damage is considered a "clean on board bill of lading," and is prima facie evidence of the good condition of the cargo at loading. See Westway Coffee Corp. v. M.V. Netuno, 675 F.2d 30, 32 (2d Cir.1982); Caemint Food, Inc. v. Brasileiro, 647 F.2d 347, 352 (2d Cir.1981); Blommer Chocolate Co. v. Nosira Sharon Ltd., 776 F. Supp. 760 (S.D.N.Y.1991). Where the defendant fails to adduce evidence tending to controvert the contents of a clean on board bill of lading, production of the bill satisfies the plaintiff's burden of proving the good condition of the cargo at loading. See Caemint Food, 647 F.2d at 352.
It is undisputed that the rail was covered with a light layer of atmospheric rust on the docks at Antwerp. It is also undisputed that the cargo was wet from freshwater and that mill scale was observable along the length of the rail. Hence, the court must address whether "rust clauses" such as those described above and appearing in the mates receipts and the two bills of lading constitute notations of damage, such that the bills of lading fail to establish the good order and condition of the rail on delivery.
Defendants cite authority for the proposition that similar rust clauses establish damage to the cargo at loading. See Acwoo Int'l v. Hosei Maru, 1989 AMC 2894, 736 F. Supp. 1452 (E.D.Mich.1989). The court disagrees. In Acwoo, the commodity at issue was a shipment of cold rolled steel. This precedent is of little value to the issue of the instant cargo, which required none of the packaging and other precautions associated with cold rolled steel and is undamaged by atmospheric rust. Moreover, as Capt. Sparks stated at trial, most cargos of hot rolled steel are shipped under bills of lading containing rust clauses similar to the one described here. Further, the definition of the relevant rust clauses in the Sparks & Co. Preshipment Survey Report recites that the insertion of such clauses does not contradict the good condition of the cargo. Since the development of surface atmospheric rust is a normal characteristic of hot rolled steel, the court finds that the instant cargo was shipped pursuant to clean bills of lading.IV.
In the alternative, defendants posit that the bills of lading do not suffice to establish the good order and condition of the cargo because the existence of mill scale on the surface of the cargo constituted an inherent vice. See The Niel Maersk, 91 F.2d 932, 934 (2d Cir.) (A. Hand, J.), cert. denied, 302 U.S. 753, 58 S. Ct. 281, 82 L. Ed. 582 (1937). When a defendant adduces evidence tending to show that damage occurred due to internal defects in the cargo, i.e., defects that are not externally observable by the carrier, the burden shifts to the shipper to disprove such a defect as part of its prima facie case. See Caemint Food, 647 F.2d at 356 n. 9 (citing Hecht, Levis & Kahn, Inc. v. The S.S. President Buchanan, 236 F.2d 627, 631 (2d Cir. 1956); American Tobacco Co. et. al. v. The Katingo Hadjipatera et. al., 81 F. Supp. 438, 446 (S.D.N.Y.1948), modified, 194 F.2d 449 (2d Cir.1951), cert. denied, 343 U.S. 978, 72 S. Ct. 1076, 96 L. Ed. 1370 (1952)).
An "inherent vice" is defined as "any existing defects, diseases, decay or the inherent nature of the commodity which will cause it to deteriorate with a lapse of time." Raphaely, supra, 972 F.2d at 504 (citing Vana Trading Co. v. S.S. Mette Skou, 556 F.2d 100, 104 (2d Cir.), cert. denied, 434 U.S. 892, 98 S. Ct. 267, 54 L. Ed. 2d 177 (1977)). In the case at bar, the testimony established that mill scale is an ordinary byproduct of the steel manufacturing process. The interaction *769 of ordinary freshwater with mill scale necessarily produces a surface layer of atmospheric rust, with corrosion cells surrounding areas of mill scale. Such rust is harmless and in no way reduces the useful life of the rail. Indeed, as already noted, the cargo in question was intended for outdoor use, where it would be permanently exposed to the elements. Additionally, both mill scale and the accompanying corrosion cells are readily observable by the naked eye and would be noticeable at loading by the carrier. Consequently, the court determines that mill scale does not constitute an inherent vice within the meaning of section 1304(2) (m), and finds that Ferrostaal has proved the good condition of the cargo at loading.V.
Nevertheless, Ferrostaal's cause of action fails on the second prong of its prima facie case. It is the essence of this case that hot rolled rail track is not considered damaged by the development of rust caused by freshwater. Thus, if the instant cargo was not damaged at loading, its discharge in substantially the same rusty condition at outturn defeats a COGSA claim. No credible scientific reports support a finding of saltwater contamination. In point of fact, the survey reports, as well as expert testimony on metallurgy and rust corrosion, established nothing more than a finding of normal atmospheric corrosion, exacerbated by heavy condensation in the cargo hold and conceivably by acid rain. More, the rejected rail was ultimately sandblasted and accepted by either PATH or NYCTA. The court agrees with defendants: whatever NYCTA communicated to Ferrostaal when it rejected the cargo, the rail was not actually damaged for NYCTA's purposes or PATH's, and was certainly not contaminated by seawater. Thus, Ferrostaal has failed to establish a prima facie case.
Since it is definitively determined that the rail was not damaged, the court does not reach the issue of the shipowner's due diligence.
The court orders that the complaint be, and it hereby is, DISMISSED. The Clerk of the Court is directed to enter judgment accordingly.
 Bernard Newman, Senior Judge of the United States Court of International Trade, sitting as a United States District Judge by designation.
 Captain Sparks is the author of Steel Carriage by Sea (1988), which is considered the authoritative work on the subject by marine surveyors. Plaintiff's Exhibit 54.
 Bill of lading # 1, numbered ALIA-AWNW90740001, covered 1008 rails, which were loaded in hold # 5. Bill of lading # 2, numbered ALIA-AWNW90740002, governed 910 rails in hold # 1. The bills of lading are dated September 16, 1989.
 During the relevant days, the logbook contained such notations as "v/l rolling & pitching violently w/rough sea & heavy swell"; "shipping seas on deck"; and "no ventilation due to heavy weather."
 The second surveyor, William A. Schell, represented the charterer's interests and was not called as a witness at trial. Schell's report and deposition testimony suggested some degree of wetting to dunnage and drip-down in the # 1 hold, but did not otherwise shed light on the condition of the cargo. Schell did not conduct silver nitrate tests, and admitted that he was not concerned with the condition of the rails. See Deposition of William Schell, Plaintiff's Exhibit 71 at 28-32. The court declines, therefore, to conclude on the basis of Schell's statements that the wetting was caused by anything other than condensation in the hold.
 Silver nitrate is a clear liquid solution used by marine surveyors for the sole purpose of detecting the presence of chlorides on the surface of cargo or within a ship's hold. The liquid is placed on an area believed to be exposed to seawater. If the liquid bubbles and turns to a milky white color, the reaction is interpreted as a positive test result. If the liquid remains clear and exhibits no reaction, the result is negative for chlorides. A positive silver nitrate result does not, however, conclusively prove that the chlorides are actually seawater; to make that determination, it is necessary to sample the exposed area and transmit the samples to a laboratory for chemical analysis.
 Farrell indicated that he obtained a "very few weak positive" results which he attributed to the saline content of the maritime atmosphere, and not to seawater.
 Two inaccuracies in the report were identified by Ferrostaal. First, a typographical error suggesting positive silver nitrate results later had to be corrected. Second, Ortiz reported seeing no streaking or drip-down on the # 1 hatch coaming, which the court determines to be in error. Nevertheless, the court considers that the balance of his report and testimony is credible, particularly in light of the consistent test results of Farrell and Tischler.
 A silver nitrate solution conforming to the Lloyd's Survey Handbook would contain 96% distilled water, 2% silver nitrate, and 2% nitric acid. Fowley admitted that he did not include nitric acid in his solution.
 The fish plate test is not actually a test for rust. Drilled rails are connected by fish plates, pieces of steel plate fitted over the ends of the rail, which are then fastened by nuts and bolts through the drilled holes. The test consists of placing a template against the side of the rail and checking to see whether the template fits within certain tolerances, thereby ensuring that the rail will not pull apart under the heavy stress of a train running over the bolted sections. Record at 23, 375.
 It is unclear whether PATH ultimately used the rail for track or guard rail. PATH did perform end hardening on the rail, which was one of the deficiencies cited by NYCTA during its commercial dispute with Ferrostaal.
 The defense of inherent vice is an excepted cause listed in section 1304(2) of the COGSA, which states in pertinent part:
(2) Uncontrollable causes of loss
Neither the carrier nor the ship shall be responsible for loss or damage arising or resulting from ...
(m) Wastage in bulk or weight or any other loss or damage arising from inherent defect, quality or vice of the goods.
46 U.S.C. § 1304(2) (m) (1988).